Research On Indoor High Precision Positioning Method Based On UWB/LiDAR Combination

In recent years,with the continuous development of IPS systems,how to obtain high-precision and stable indoor positioning data has become a hot topic of concern to all parties.Different from the outdoor environment,which adopts more mature systems including GPS,the positioning accuracy can meet the requirements for positioning data in people's lives.There are more flows and fixed obstacles that cause NLOS errors in the indoor environment.Using a single sensor cannot guarantee the accuracy and continuity of positioning.Therefore,sensor data fusion has become a powerful means to improve positioning accuracy.Based on UWB and LiDAR,this paper analyzes the positioning methods,error sources and data characteristics of the both in indoor scenes,respectively performs data preprocessing,then tries to adopt a certain mutual constraint strategy to fuse the two datas.By using LiDAR to provide more accurate pose changes,as well as UWB's ultra-high time resolution and good anti-multipath effects,the positioning capability and accuracy of the combined system are more robust than a single system.The main tasks completed in this paper include:(1)An overview of the LiDAR system,in view of the existence of motion distortion in the original LiDAR data,the assumption based on the uniform motion model is used to remove;after that,two commonly used LiDAR point cloud data registration methods are compared and analyzed,and the main model implemented is derived and carried out.According to the experimental analysis,the method based on incremental NDT is used to process point cloud data with higher accuracy and efficiency.(2)Several positioning modes in the UWB system are explained,and the error analysis of the acquired data is carried out.The fourth-order polynomial is used to correct the standard time deviation and the wavelet transform is used to remove the UWB NLOS error.Experiments show that this method can obtain relatively smooth UWB ranging information.(3)It is difficult to use a single sensor to solve the positioning problem in a complex indoor environment.Proposes a positioning model based on EKF-based UWB/LiDAR odometry,adopts the constraint strategy of angle information,and derives the realization process in detail,realizes the unification of time and space,effectively integrates UWB and LiDAR data and complements each other.(4)In this paper,a UWB/LiDAR combined mobile positioning platform is built,and experiments are carried out in indoor meeting rooms and underground parking lots in complex environments.The experimental results show that the positioning accuracy can reach the decimeter level,and the positioning accuracy is increased by 35% and 60%,respectively.The paper has 51 figures,7 tables and 64 references.